The influence of encapsulating carbohydrates (EC) with varying properties on the technological and functional properties of jussara pulp microparticles produced by spray drying were evaluated using experimental design.
Carbohydrate Polymers 151 (2016) 500–510 Contents lists available at ScienceDirect Carbohydrate Polymers journal homepage: www.elsevier.com/locate/carbpol Starch, inulin and maltodextrin as encapsulating agents affect the quality and stability of jussara pulp microparticles Ellen Cristina Quirino Lacerda a , Verônica Maria de Araújo Calado b , Mariana Monteiro c , Priscilla Vanessa Finotelli d , Alexandre Guedes Torres a , Daniel Perrone a,∗ a Laboratório de Bioqmica Nutricional e de Alimentos, Departamento de Bioquímica, Instituto de Química, Universidade Federal Rio de Janeiro, Av Athos da Silveira Ramos, 149, CT, Bloco A, sala 528A, 21941-909, Rio de Janeiro, Brazil b Escola de Química, Universidade Federal Rio de Janeiro, Av Athos da Silveira Ramos, 149, CT, Bloco E, 21941-909, Rio de Janeiro, Brazil c Laboratório de Alimentos Funcionais, Instituto de Nutric¸ão Jos de Castro, Universidade Federal Rio de Janeiro, Av Carlos Chagas Filho, 373, CCS, Bloco J, 2◦ andar, sala 16, 21941-902, Rio de Janeiro, Brazil d Laboratório de Nanotecnologia Biofuncional, Faculdade de Farmácia, Universidade Federal Rio de Janeiro, Av Brigadeiro Trompowski, s/n◦ , CCS, Bloco A2, sala 38, 21941-590, Rio de Janeiro, Brazil a r t i c l e i n f o Article history: Received 19 November 2015 Received in revised form 24 May 2016 Accepted 25 May 2016 Available online 27 May 2016 Chemical compounds studied in this article: Cyanidin-3-O-glucoside (PubChem CID: 441667) Cyanidin-3-O-rutinoside (PubChem CID: 441674) Keywords: Anthocyanins Antioxidant activity Instrumental color Natural colorant Shelf-life stability Spray drying a b s t r a c t The influence of encapsulating carbohydrates (EC) with varying properties on the technological and functional properties of jussara pulp microparticles produced by spray drying were evaluated using experimental design Microparticles produced with sodium octenyl succinate (OSA) starch at 0.5 core to EC ratio and with mixtures of inulin and maltodextrin at 1.0 and 2.0 core to EC ratio showed darker color, and higher anthocyanins contents and antioxidant activity Seven microparticles showing high water solubility and desirable surface morphology Hygroscopicity (10.7% and 11.5%) and wettability (41 s and 43 s) were improved when OSA starch and mixtures of inulin and maltodextrin were used The anthocyanins contents and color of the microparticles did not change when exposed to light at 50 ◦ C for 38 days Finally, microparticles produced at 1.0 core to EC ratio with 2/3 OSA starch, 1/6 inulin and 1/6 maltodextrin were selected These microparticles may be applied as colorant in numerous foods, whilst adding prebiotic fiber and anthocyanins © 2016 Elsevier Ltd All rights reserved Introduction Jussara (Euterpe edulis Martius) is a palm tree native of the Brazilian Atlantic Forest from which a noble type of palm heart is produced (Borges et al., 2011) However, its intense and unsustainable exploitation in the last few decades made it an endangered species, being found only in natural reserves (Silva, Carmo et al., 2013) Therefore, the economic valorization of jussara palm fruits, which are exclusively obtained by extractivism, instead of the palm heart, seems to be a sustainable way for its management as well ∗ Corresponding author E-mail addresses: ellenelle1@gmail.com (E.C.Q Lacerda), calado@eq.ufrj.br (V.M.d.A Calado), mariana@nutricao.ufrj.br (M Monteiro), pfinotelli@gmail.com (P.V Finotelli), torres@iq.ufrj.br (A.G Torres), danielperrone@iq.ufrj.br, danielperronemoreira@gmail.com (D Perrone) http://dx.doi.org/10.1016/j.carbpol.2016.05.093 0144-8617/© 2016 Elsevier Ltd All rights reserved as the Atlantic Forest However, jussara palm fruits are relatively unknown to the market and consumers, partly because they are highly perishable Jussara palm fruit is a berry composed by a single light brown seed covered by a dark purple thin and dry skin that represents approximately 10% of the whole fruit weight (Borges et al., 2011; Inada et al., 2015) The fruit is usually added to warm water to separate it from the seeds by softening the skin, to yielding a very thick dark purple juice (Supplementary Fig S1) The purple color of jussara pulp is attributed to the anthocyanins, mainly cyanidin-3O-glucoside and cyanidin-3-O-rutinoside (Inada et al., 2015) that can be used as food colorants (He & Giusti, 2010) These compounds are associated with potential health benefits in humans (Nile & Park, 2014) Although jussara is an interesting source of these pigments, the fruit is highly perishable and anthocyanins are susceptible to degradation when isolated (He & Giusti, 2010) E.C.Q Lacerda et al / Carbohydrate Polymers 151 (2016) 500–510 Spray drying microencapsulation of solids, liquids or gaseous materials with thin polymeric coatings is a well-established technology (Gharsallaoui, Roudaut, Chambin, Voilley & Saurel, 2007), which could be employed for the stabilization of anthocyanins in jussara fruit, as well as increasing its shelf-life (Ersus & Yurdagel, 2007) Different types of encapsulating agents, such carbohydrates and proteins may be used alone or combined (Gharsallaoui et al., 2007) Carbohydrates such as gums (Bicudo et al., 2015), starch (Tonon, Brabet & Hubinger, 2010), modified starches (Villacrez, Carriazo & Osorio, 2014), dextrins (Ersus & Yurdagel, 2007) and cellulose (Yousefi, Emam-Djomeha, Mousavi, Kobarfard & Zbicinski, 2015) are vastly employed and act as a protective film, isolating the core and avoiding the deleterious effects of external factors (Silva, Stringheta, Teófilo & Oliveira, 2013) The choice of the carbohydrate used as wall material should take into consideration its cost, properties related to the food application, such as solubility or emulsifying activity, as well as added value to the product, such as bioactivity Maltodextrin is obtained by enzymatic or acid hydrolysis of starch and generally shows high water solubility, low viscosity and low cost (Gibbs, Kermasha, Alli & Mulligan, 1999) Its low emulsifying capacity can be overcome by the use of modified starches Starch sodium octenyl succinate (OSA starch) is a chemically modified starch that contains a lipophilic component that enhances its amphiphilic character and thus its emulsifying capacity (Sweedman, Tizzotti, Schäfer, & Gilbert, 2013) Another interesting carbohydrate for encapsulation is inulin, a polysaccharide composed of fructose units linked by -(2,1) bonds and containing a glucose unit Inulin can be commercially obtained from chicory and has prebiotic effects, dietary fiber actions, among other health related benefits (Ranawana, 2008) The aim of this work was to develop a natural food colorant with attractive and stable color characteristics that is also rich in potentially bioactive anthocyanins This product would add economic value to jussara and hopefully aid in the preservation of the Brazilian Atlantic Forest The present study evaluates the effects of carbohydrate sources with varying properties used individually or in combinations as encapsulating agents on the quality and stability of jussara pulp microparticles obtained by spray drying Material and methods 2.1 Standards and chemicals 2,4,6-Tris-(2-pyridyl)-S-triazine (TPTZ), 2,2 -azino-bisacid) diammonium (2-ethylbenzothiazoline-6-sulfonic salt (ABTS), potassium persulfate, (±)-6-hydroxy2,5,7,8-tetramethylchromane-2-carboxylic acid (Trolox), 2,2 -azobis(2-methylpropionamidine) dihydrochloride (AAPH) and fluorescein were purchased from Sigma-Aldrich Chemical Co (St Louis, MO, USA) Iron (II) sulfate was purchased from Merck KGaA (Darmstadt, HE, Germany) Cyanidin-3-O-glucoside and cyanidin-3-O-rutinoside standards were purchased from Indofine® Chemical Company (Hillsborough, NJ, USA) OSA starch (commercial brand Capsul® ) with a degree of substitution of 0.0199, emulsifying activity index of 0.092 and emulsion stability index of 0.984 was purchased from National Starch (Bridgewater, NJ, USA) Maltodextrin MOR-REX 1920 with 22 Dextrose Equivalents (DE) was purchased from Corn Products (Mogi Guac¸u, Brazil) Inulin with a degree of polymerization of 10 (91% fructose and 9% glucose) was purchased from Siba Ingredients (São Paulo, Brazil) The FTIR spectra (Supplementary Fig S2) and the thermogravimetric curves (Supplementary Fig S3) of OSA starch, inulin and maltodextrin were consistent with literature data The succinyl moiety of OSA modified starch was confirmed due to the presence 501 of a carbonyl group signal by FTIR spectroscopy (Supplementary Fig S2) All solvents were of HPLC grade from Tedia (São Paulo, Brazil) HPLC grade water (Milli-Q system, Millipore, Bedford, MA, USA) was used throughout the experiments 2.2 Preparation of jussara pulp microparticles by spray drying Frozen jussara palm (E edulis) fruit pulp was donated by Jucáaớđ processing company, located in Resende (Rio de Janeiro State, Brazil) Fruits (2.5 kg) were immersed in water (1 L) at 40 ◦ C for 30 for peel softening, processed in a vertical depulper to yield jussara pulp and stored at −20 ◦ C Prior to the preparation of microparticles, jussara pulp was thawed and centrifuged (1640g, 10 min, 25 ◦ C) (SorvallTM ST 16 R centrifuge, Thermo Fisher Scientific Inc, USA) The spray drying feed solution was prepared with this supernatant in order to remove suspended solids and facilitate the product passage through the nozzle atomizer The encapsulating carbohydrates (EC) were added to the centrifuged pulp and stirred with a magnetic bar until complete dissolution Finally, the mixture was sonicated at 90% amplitude for using UP100H ultrasonic processor (Hielscher Ultrasonics, Teltow, Germany) Jussara pulp microparticles were produced according to a simplex-lattice design with 3factors (OSA starch, inulin and maltodextrin as EC), interior points and central point with two repetitions, for each core (jussara pulp) to EC ratio (0.5, 1.0 and 2.0, w/w), totaling 33 runs (Table 1) The core to EC ratios were calculated according to the total solids content of the centrifuged pulp Table Experimental design for optimization of jussara pulp microparticles production with different encapsulating carbohydrates (EC) and core to EC ratios, with their corresponding encapsulating process yield Run Encapsulating Carbohydrates (EC) OSA starch 10b 11b 12 13 14 15 16 17 18 19 20 21b 22b 23 24 25 26 27 28 29 30 31 32b 33b 0 1/2 1/2 2/3 1/6 1/6 1/3 1/3 0 1/2 1/2 2/3 1/6 1/6 1/3 1/3 0 1/2 1/2 2/3 1/6 1/6 1/3 1/3 Inulin 1/2 1/2 1/6 2/3 1/6 1/3 1/3 1/2 1/2 1/6 2/3 1/6 1/3 1/3 1/2 1/2 1/6 2/3 1/6 1/3 1/3 Core to EC ratios Process Yield (%)a 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0 2.0 51.5 44.2 43.5 51.3 53.1 41.6 61.1 47.3 31.9 46.6 44.8 55.8 38.9 40.6 51.4 52.3 21.5 33.4 29.4 40.9 49.8 40.6 49.3 22.4 38.8 33.4 42.0 39.8 43.8 22.6 29.7 32.0 28.9 Maltodextrin 0 1/2 1/2 1/6 1/6 2/3 1/3 1/3 0 1/2 1/2 1/6 1/6 2/3 1/3 1/3 0 1/2 1/2 1/6 1/6 2/3 1/3 1/3 a Calculated as the ratio between the mass of microparticles obtained and the mass of total solids in the feed solution b Central point 502 E.C.Q Lacerda et al / Carbohydrate Polymers 151 (2016) 500–510 (5.5g/100 g), which was determined gravimetrically using an MA35 infrared moisture analyzer (Sartorius, Goettingen, Germany) Spray drying was performed on a Mini Spray Dryer Büchi 290 (Büchi Laboratoriums Technik, Flawil, Switzerland), coupled with a 0.3 mm diameter nozzle and inlet air temperature set at 140 ± ◦ C for all experiments (Tonon, Brabet & Hubinger, 2008) The outlet temperature was not controlled (59.6 ◦ C on average) The feed solution was kept under magnetic stirring and pumped into the main drying chamber by a peristaltic pump, with aspiration rate of 32 m3 /h, compressor air pressure of 0.03 MPa and feed flow rate of 0.36 L/h Spray drying process yield was calculated as the amount of particles obtained in both the collector and cyclone in relation to the total solids present in the feed solution (Table 1) Esteve & Frígola, 2009; Benzie & Strain, 1996; Re et al., 1999) ORAC, FRAP and TEAC values were expressed as, respectively, mmol of Trolox equivalents per g, mmol of Fe2+ equivalents per g and mmol of Trolox equivalents per g Each extract was analyzed in triplicate 2.3 Instrumental color 2.9 Water solubility, hygroscopicity and wettability Instrumental color of jussara pulp microparticles was determined using Minolta CR-400 colorimeter (Konica Minolta, Osaka, Japan), with illuminant D65, 2◦ viewing angle The CIELab color space was used to determine the color components: L* [black (0) to white (100)], a* [greenness (−) to redness (+)] and b* [blueness (−) to yellowness (+)] The coordinates L*a*b* were measured directly from the dry microparticles in triplicate For the stability test, the total color difference ( E*) was calculated according to Eq (1) Water solubility of selected jussara pulp microparticles was determined as described by Cano-Chauca, Stringheta, Ramos and Cal-Vidal (2005) with modifications Microparticles (100 mg) were dissolved in water (10 mL), centrifuged and the amount of soluble solids in the supernatant was determined gravimetrically and related to total solids in microparticles Hygroscopicity was determined by storing selected jussara pulp microparticles (500 mg) in a desiccator with saturated NaCl solution (75.0% relative humidity) at room temperature for days After this period, microparticles were weighed and hygroscopicity was expressed as percentage (Tonon, Brabet, Pallet, Brat & Hubinger, 2009a) Wettability of selected jussara pulp microparticles was determined according to the method proposed by Hla and Hogekamp (1999) with adaptations Microparticles (250 mg) were added, without agitation, on water (50 mL) at 25 ◦ C in an apparatus with 47.5 cm2 of surface area Wettability was defined as the time in seconds required for microparticles to be visually completely wet 2.10 Stability during storage at 50 ◦ C The stability during storage of microparticles selected from the experimental design according to desired technological, functional and physicochemical characteristics was investigated Jussara pulp microparticles were placed in transparent polyester bags and vacuum-sealed To investigate the effect of light exposure, half of the bags were covered with aluminum foil to protect it against light Then, samples were placed in an oven at 50 ◦ C and approximately relative humidity of 50% with constant exposure to fluorescent light Protected and unprotected microparticles were analyzed at days intervals for up 38 days for color (according to section 2.3), anthocyanins content (according to section 2.6) and antioxidant activity by FRAP and TEAC assays (according to section 2.7) E∗ = Lt∗ − Lt∗ i + a∗t − a∗t i + b∗t − b∗t i (1) where L*, a* and b* are the color coordinates at initial time (t0 ) and at nth day of storage (ti ) time 2.4 Water activity Water activity (aw ) of jussara pulp microparticles was measured directly at 25 ◦ C using the LabMaster-aw analyzer (Novasina, Pfáffikon, Switzerland) 2.5 Particle size distribution Particle size distribution was measured using the SALD-2201 laser diffraction particle size analyzer (Shimadzu, Tokyo, Japan), with measuring range from 0.3 to 1000 m Jussara pulp microparticles were suspended in isopropyl alcohol, sonicated with a microtip probe, and particle size distribution was determined after successive readings became constant Particles size dispersion was defined as the coefficient of variation of the particle size distribution analysis 2.6 Anthocyanins contents and retention Jussara pulp microparticles were completely dissolved in water (0.2%, w/v) into an ultrasound bath Anthocyanins were analyzed according to Inada et al (2015) on a liquid chromatography system (Shimadzu® ), which included a quaternary pump LC-20AT, automatic injector SIL-20AHT, diode-array detector (DAD) SPD-M20A, system controller CBM-20A and degasser DGU-20A5 Anthocyanins retention was calculated using anthocyanin content found in the feed solution (considering its content as 2.59 mg/g according to previous HPLC analysis) and anthocyanin content in microparticles 2.7 Antioxidant activity The antioxidant activity (AA) was determined in the same extracts used for anthocyanin analysis using ORAC (Oxygen Radical Absorbance Capacity), FRAP (Ferric Reducing Antioxidant Power) and TEAC (Trolox Equivalent Antioxidant Capacity) assays (Zulueta, 2.8 Particles morphology Selected jussara pulp microparticles were deposited on carbon double-sided adhesive tape mounted on stubs, coated with gold under vacuum The stubs were observed in a scanning electron microscope (JEOL® , JSM-6460 LV, Tokyo, Japan), operated at 20 kV with a magnification of 2000 2.10 Statistical analysis The experimental matrix was generated by the Statistica software, version 7.0 (StatSoft Inc., Tulsa, OK, EUA) The effect of EC composition was evaluated for each core to EC ratio by analysis of variance (ANOVA) using Statistica software Water solubility, hygroscopicity and wettability of selected jussara pulp microparticles were compared by ANOVA followed by Tukey’s multiple comparison post-test using GraphPad Prism software Windows, version 5.04 (GraphPad Software, San Diego, CA, USA) Stability data was analyzed by two-way ANOVA followed by Tukey’s multiple comparison post-test using GraphPad Prism software The effect of time of storage was evaluated by comparing microparticles at initial (t = 0) and final times (t = 38 days), while the effect of light exposure was evaluated by comparing microparticles protected and unprotected from light at the final time (t = 38 days) Results were considered significant when p < 0.05 E.C.Q Lacerda et al / Carbohydrate Polymers 151 (2016) 500–510 Results and discussion 3.1 Effect of encapsulating carbohydrates composition on jussara pulp microencapsulation The effect of different EC mixtures using different core to EC ratios on technological (instrumental color, aw , particles size dispersion) and functional properties (anthocyanins contents and retention, and AA) of jussara pulp microparticles was assessed using a simplex-lattice experimental design (Table 2) 3.1.1 Technological properties Color, aw and size dispersion are technological properties of great importance when developing powders intended for employment as food colorants aw is related to shelf life of the product, as it affects microbial growth and enzymatic and non-enzymatic degradation reactions The size dispersion of the microparticles is related to physicochemical properties that influence water solubility, hygroscopicity and wettability The effect of EC composition, for each core to EC ratio, on L* and a* color coordinates, aw and particles size dispersion are shown in Fig Although jussara pulp microparticles were visually purple (Supplementary Fig S4), only L* and a* coordinates were considered when choosing the EC composition, as b* coordinate presented values close to zero Therefore, desirable microparticles would present low L* values and high positive a* values, which correspond to dark microparticles of predominantly red hue L* values of jussara pulp microparticles ranged from 23.9 (run #28) to 46.5 (run #4) (Table 2; Fig 1); and core to EC ratios rather than EC composition was the main factor that affected this color component We observed that the higher the core to EC ratio, the darker were the microparticles produced, which may be explained by the light color of EC opposed to the dark color of jussara pulp Moreover, this behavior could also be attributed to the formation of dark pigments due to caramelization of sugars (fructose and glucose) present in jussara pulp (Inada et al., 2015) This hypothesis is supported by the formation of sticky particles that adhered to the inner chamber of the spray dryer when higher core to EC ratios were employed for microparticles production Jussara pulp microparticles showed a predominantly red hue, with a* values ranging from 10.9 (run #30) to 25.2 (run #2) and b* values ranging from −0.8 (run #2) to 3.9 (run #11) (Table 2) In general, lower core to EC ratios led to slightly more intense red hue microparticles Independently of core to EC ratio, microparticles produced with EC containing high proportions of OSA starch presented higher a* values (Fig 1) Blackberry microparticles produced with maltodextrin, gum arabic and their mixture showed L* (36.0–39.9), a* (19.1–23.4) and b* (3.6–3.9) values (Ferrari, Germer, Alvim, Vissotto & Aguirre, 2012) similar to those observed for jussara pulp microparticles produced with 0.5 and 1.0 core to EC ratios Ac¸ pulp microparticles produced with maltodextrin 10DE at 0.5 core to EC ratio showed lighter color (L* = 54.5, a* = 10.8 and b* = 2.4) (Tonon, Brabet & Hubinger, 2009b) than the corresponding condition in the present study (run #3, L* = 37.2, a* = 20.6 and b* = 1.2) All jussara pulp microparticles presented low aw values, ranging from 0.252 (run #7) to 0.484 (run #25) (Table 2), considered to be adequate for this type of product as well as inhibiting microbial growth Moreover, the relative rate of oxidative reactions is low at aw values close to 0.3, which were observed for most microparticles produced with lower core to EC ratios, suggesting that the color of these products would be stable Higher aw values (0.40–0.44) were observed for jussara microparticles produced with gelatin, maltodextrin 20 DE and gum arabic at 165 ◦ C and 1.0 core to EC ratio (Bicudo et al., 2015) Nevertheless, we observed that microparticles produced with higher core to EC ratio presented higher aw values, which may be explained by water retention in the core due 503 to higher relative amounts of encapsulated jussara pulp, which contains hygroscopic components, such as simple sugars (Inada et al., 2015) This hypothesis is corroborated by the inverse correlation between aw and L* values (r = −0.691, p < 0.0001, n = 33) Jussara pulp microparticles showed mean diameter values ranging from 1.84 m to 12.08 m (Table 2), slightly smaller than that of ac¸ pulp microparticles (9.01 m) produced with maltodextrin 20 DE by spray drying at 140 ◦ C (Tonon et al., 2010) In powder foods, particle size dispersion is usually more important than mean diameter, as it influences aspects of processing, handling and shelf life (Tonon, Brabet, Pallet et al., 2009) Narrower particle size dispersion represents more homogenous physicochemical properties related to the employment of food powders, such as water solubility, wettability and hygroscopicity In the present study, particle size dispersion was evaluated by the coefficient of variation of the mean diameter of the microparticles and ranged from 4.6% (run #1) to 30.2% (run #22) (Table 2) Mean particle diameter and size dispersion were inversely correlated (r = −0.787, p < 0.0001, n = 33), indicating that conditions that led to bigger microparticles should be preferred over smaller ones In general, higher core to EC ratios led to particles with higher mean particle diameter and lower size dispersion, possibly due to the aggregation of the sugars present in the pulp On the other hand, at 0.5 core to EC ratio, the EC composition had limited influence on mean diameter and size dispersion Considering that water solubility and viscosity of the encapsulating material are related to mean particle diameter (Elversson & Millqvist-Fureby, 2005) and therefore size dispersion, carbohydrates with low solubility and high viscosity, such as OSA starch, should be favored over carbohydrates with high solubility and low viscosity, such as maltodextrin, for producing microparticles with desirable functional properties 3.1.2 Functional properties Anthocyanins contents ranged from 3.3 mg/g (run #4) to 24.2 mg/g (run #1) (Table 2) Tonon et al (2010) obtained ac¸ pulp microparticles using maltodextrin 20 DE at 0.5 core to EC ratio with a higher anthocyanins content (34.0 mg/g) than the corresponding condition in the present study (run #3, 9.9 mg/g) On the other hand, Ersus and Yurdagel (2007) obtained black carrots microparticles under similar conditions with anthocyanins contents of 6.3 mg/g These differences may be explained by the concentration of anthocyanins in the core materials, spray drying conditions, as well as the analytical procedure used to quantify anthocyanins Authors who have studied anthocyanins encapsulation usually employ the differential pH colorimetric assay, while we used HPLC analysis in the present study Similarly to jussara pulp (Borges et al., 2011; Inada et al., 2015), cyanidin-3-O-rutinoside and cyanidin-3-O-glucoside were the two anthocyanins identified in microparticles and corresponded to 24% and 76%, respectively, of total anthocyanins As expected, anthocyanins contents were inversely correlated with L* values (r = −0.488, p = 0.004, n = 33) Surprisingly, no correlations were found between anthocyanins contents and both a* and b* values, as one would expect based on the characteristic reddish-purple color of cyanidins Nevertheless, when considering only microparticles produced with 0.5 core to EC ratio, anthocyanins contents were correlated to both a* (r = 0.796, p = 0.003, n = 11) and b* values (r = −0.676, p = 0.022, n = 11) Jussara pulp microparticles produced with 1.0 and 2.0 core to EC ratio presented a darker color, which may have compromised the colorimeter ability to distinguish small differences in red and blue hues of these samples Anthocyanins retention in microparticles ranged from 6.0% (run #30) to 67.0% (run #1) (Table 2) and was correlated to both anthocyanin contents (r = 0.620, p = 0.0001, n = 33) and encapsulating process yields (r = 0.441, p = 0.0102, n = 33) (Table 1) Under similar drying conditions, higher retention values were reported for 504 E.C.Q Lacerda et al / Carbohydrate Polymers 151 (2016) 500–510 Table Instrumental color, water activity (aw ), particles mean size and dispersion, anthocyanins contents and retention, and antioxidant activity of jussara pulp microparticles produced according to the experimental design of Table 1.a Run 10h 11h 12 13 14 15 16 17 18 19 20 21h 22h 23 24 25 26 27 28 29 30 31 32h 33h Instrumental Colorb L* a* b* 32.5 38.8 37.2 46.5 37.6 41.6 40.5 35.1 32.1 43.1 45.8 34.6 30.9 29.7 35.7 37.5 27.5 31.3 30.9 30.2 39.0 38.6 28.2 25.7 25.4 32.1 31.7 23.9 25.1 24.4 25.1 24.3 23.9 24.2 25.2 20.6 17.2 23.4 22.7 23.5 23.8 23.7 19.9 19.6 23.1 21.2 18.7 22.7 22.4 17.0 23.1 21.0 17.6 21.7 22.7 21.4 19.9 16.4 20.8 20.9 15.3 14.6 10.9 15.0 13.9 14.4 0.1 −0.8 1.2 3.3 0.6 −0.3 0.6 0.4 0.4 1.9 3.9 1.4 2.2 1.1 1.2 1.4 2.1 0.6 0.7 1.7 1.2 1.8 0.7 2.3 1.8 2.2 1.7 2.3 1.8 2.2 1.6 2.1 1.8 aw 0.331 0.351 0.293 0.322 0.393 0.339 0.252 0.330 0.301 0.343 0.346 0.353 0.327 0.368 0.339 0.339 0.434 0.464 0.434 0.358 0.310 0.306 0.433 0.457 0.484 0.390 0.360 0.442 0.338 0.428 0.449 0.422 0.444 Particles Size Anthocyanins Antioxidant Activity Mean (m) Dispersion (%)c Content (mg/g) Retention(%)d ORACe FRAPf TEACg 9.8 4.7 5.3 4.6 3.9 4.5 4.6 4.1 2.8 3.4 3.9 3.7 4.1 4.2 3.4 8.6 3.3 4.5 8.1 6.3 1.9 1.8 8.7 9.2 4.3 1.8 10.5 11.4 8.9 12.1 8.9 2.0 1.8 4.6 11.3 8.6 11.3 13.7 11.7 11.3 11.4 17.3 16.1 12.4 14.8 14.9 12.6 16.0 5.4 14.4 13.7 6.0 8.2 28.5 30.2 5.9 6.0 12.2 28.0 5.4 5.3 5.9 5.0 7.3 12.0 11.4 24.2 14.2 9.9 3.3 13.7 11.5 10.9 13.9 21.1 5.0 7.3 10.3 7.96 14.4 8.32 4.82 10.2 16.2 15.8 7.00 8.21 8.78 14.8 16.8 14.0 7.53 8.98 18.7 14.7 9.81 16.5 12.2 11.3 67.0 33.7 23.2 9.1 39.1 25.7 35.8 35.3 36.1 12.6 21.7 20.6 11.1 20.9 15.3 9.0 7.8 19.4 16.6 10.3 14.7 12.8 19.7 10.1 14.6 6.8 10.1 20.0 17.3 6.0 13.2 9.7 12.5 839.0 706.4 486.2 530.0 359.8 432.0 484.4 452.4 520.0 482.5 652.7 340.8 314.0 299.1 266.6 350.6 288.0 280.2 321.6 270.4 233.7 192.9 657.9 395.4 299.7 411.7 324.4 502.4 346.1 340.0 462.6 411.5 341.7 227.0 181.2 161.8 138.6 163.6 167.5 167.4 176.9 218.8 144.4 132.0 139.9 138.7 166.9 138.1 153.7 149.4 171.7 193.9 164.8 141.2 141.1 182.0 186.5 170.9 148.2 146.6 190.6 155.7 126.3 208.3 153.9 166.9 989.9 842.3 750.6 733.0 840.6 862.2 811.0 833.7 988.8 615.0 600.9 596.6 585.4 577.7 601.9 625.1 596.1 670.1 631.9 635.6 630.9 584.7 924.5 992.9 827.5 574.3 600.4 850.6 718.8 555.8 967.6 807.8 698.5 a Antioxidant activity analyses were performed in triplicate, anthocyanins contents, instrumental color and particles size analyses were performed in duplicate and water activity analysis was performed as a single replicate Analyses were performed at room temperature (28 ± ◦ C) All results presented a coefficient of variation lower than 10% b CIELab color space was used to determine coordinates L* [black (0) to white (100)], a* [green (–) to red (+)] and b* [blue (–) to yellow (+)] c Calculated as the coefficient of variation of the mean particle size d Calculated as the amount of anthocyanins in microparticles in relation to that in the feed solution e ORAC (Oxygen Radical Absorbance Capacity) expressed in mmol Trolox/g f FRAP (Ferric Reducing Antioxidant Power) expressed as mmol Fe+2 /g g TEAC (Trolox Equivalent Antioxidant Capacity) expressed as mmol Trolox/g h Central point microencapsulation of ac¸ pulp with maltodextrin (82% on average) (Tonon et al., 2008), jabuticaba extract with maltodextrin and maltodextrin/OSA starch (83% on average) (Silva, Stringheta et al., 2013), and jussara pulp with maltodextrin (75%) (Bicudo et al., 2015) In the present study, retention was compromised by the low encapsulating process yields (Table 1), especially for microparticles produced with 2.0 core to EC ratio (22% to 49%), which adhered inside the spray dryer chamber Similar process yields were observed for encapsulation of a M citrifolia L fruit extract with maltodextrin using 0.5 (39.2%) and 1.0 (20.7%) core to EC ratios (Krishnaiah, Sarbatly & Nithyanandam, 2012) In general, spray drying with 50% recovery in the cyclone is considered as successful (Sahin-Nadeem & Özen, 2014) Another hypothesis that might explain the low retention values observed in our study is that we quantified cyanidin-3-O-rutinoside and cyanidin-3-O-glucoside by HPLC while those other authors determined total anthocyanins by the differential pH assay Therefore, heat-induced chemical transformations in specific anthocyanins, such as isomerization and loss of sugar moiety, may have not been detected by this non-specific colorimetric assay ORAC, FRAP and TEAC values were correlated with each other (r > 0.470, p < 0.005, n = 33) and ranged, respectively, from 192.9 mmol Trolox/g (run #22) to 839.0 mmol Trolox/g (run #1), 126.3 mmol Fe+2 /g (run #30) to 227.0 mmol Fe+2 /g (run #1), and from 555.8 mmol Trolox/g (run #30) to 992.9 mmol Trolox/g (run #24) (Table 2) Anthocyanins contents were strongly correlated with FRAP (r = 0.864, p < 0.0001, n = 33) and TEAC values (r = 0.711, p < 0.0001, n = 33) Only a weak correlation was observed between anthocyanins contents and ORAC values (r = 0.370, p = 0.0341, n = 33), possibly due to interference of the EC composition in this assay Inulin presents superoxide-radical and hydroxyl radical scavenging activities (Ren, Liu, Dong & Guo, 2011), which follow a mechanism similar to ORAC assay In fact, by removing from the dataset microparticles produced with higher proportions of inulin, a much stronger correlation was observed between anthocyanins contents and ORAC values (r = 0.611, p = 0.0156, n = 15) The effect of EC composition, for each core to EC ratio, on anthocyanins contents and AA are shown in Fig At 0.5 core to EC ratio, higher proportions of OSA starch improved these variables At higher core to EC ratios, microparticles produced with mixtures of maltodextrin and inulin showed higher anthocyanins contents and AA This behavior may be related to the relative polarity of these carbohydrates When higher core to EC ratios were used, the hydrophilic anthocyanins present in jussara pulp were more properly encapsulated by the more hydrophilic EC, maltodextrin and inulin On the other hand, OSA starch favored encapsulation E.C.Q Lacerda et al / Carbohydrate Polymers 151 (2016) 500–510 505 Fig Instrumental color (coordinates L* and a*), water activity (aw ) and size dispersion fitted surfaces of jussara pulp microparticles produced with different encapsulating carbohydrates (EC: OSA starch, inulin and maltodextrin) and core to EC ratios when a lower core to EC ratio was used This may be explained by the lipophilic succinyl moiety present in OSA starch structure that would simultaneously interact with polar and non-polar molecules, allowing the encapsulation of jussara pulp as a whole, including its lipids and other lipophilic components (Inada et al., 2015) To select the microparticles for further investigations on their physicochemical attributes, data obtained for all 33 runs of the experimental design were considered Firstly, we prioritized microparticles with high anthocyanins contents and dark color (low L* values) From this pre-selection, we narrowed our choice to those 506 E.C.Q Lacerda et al / Carbohydrate Polymers 151 (2016) 500–510 Fig Anthocyanins contents and antioxidant activity (ORAC, FRAP and TEAC assays) fitted surfaces of jussara pulp microparticles produced with different encapsulating carbohydrates (EC: OSA starch, inulin and maltodextrin) and core to EC ratios microparticles that showed low values of particle size dispersion and high antioxidant activity Water activity was not considered as all microparticles showed desirable aw values By using these criteria, seven runs were selected (#1, #9, #18, #19, #24, #28, #31) (Table 1) It is worth noting that the selected microparticles were produced with all three core to EC ratios studied In addition, for each of this core to EC ratios, a different EC composition was used, with the exception of runs #9 and #31 (1/6 OSA starch, 1/6 inulin and 2/3 maltodextrin) E.C.Q Lacerda et al / Carbohydrate Polymers 151 (2016) 500–510 3.2 Physicochemical characterization of selected jussara pulp microparticles Water solubility ranged from 76.8% to 85.0% (Table 3), indicating that all selected microparticles were adequately soluble in water Similar results were obtained for water solubility of black mulberry powders produced with maltodextrin (from 77% to 85%) (Fazaeli, Emam-Djomeh, Ashtari & Omid, 2012) Water solubility was not affected by core to EC ratio or EC composition, which may be explained by both the low content of insoluble solids in centrifuged jussara pulp and the high solubility of the carbohydrates employed in the present study Hygroscopicity values ranged from 10.7% to 14.3% (Table 3), indicating that jussara pulp microparticles may be considered as slightly hygroscopic, preventing solubilization and softening of the encapsulating agent and consequent exposure of the core to atmospheric oxygen (Silva, Stringheta et al., 2013) Tonon, Brabet, Pallet et al (2009) observed higher hygroscopicity values for ac¸ pulp microparticles produced with maltodextrin 20 DE (19.7%) EC composition affected the hygroscopicity of jussara microparticles, with the lowest values observed for runs #1 (10.7%) and #18 (11.5%), in which higher proportions of OSA starch were used In contrast, the highest hygroscopicity were observed for runs #31 (14.3%), #19 (14.1%) and #24 (13.7%), for which higher core to EC ratios and higher proportions of maltodextrin and specially inulin were used (Table 3) These results may be explained by differences in chemical structure of the carbohydrates used in the study While OSA starch has an amphiphilic character, showing lower water absorbing ability (Silva, Stringheta et al., 2013), inulin is considered as highly hygroscopic due to its branched structure, which facilitates hydrogen bonding and thus moisture absorption from ambient air (Akalin & Eris¸ir, 2008) Similar, to our results, jabuticaba peel extract (Silva, Stringheta et al., 2013) and propolis (Silva, Fonseca et al., 2013) microparticles produced with OSA starch by spray drying showed lower hygroscopicity than those produced with gum arabic and maltodextrin Moreover, when microparticles were produced at 0.5 or 1.0 core to EC ratio, the EC probably hurdled the highly hygroscopic sugars present in jussara pulp (Phisut, 2012) Faster wettability is a desirable feature for the development of instant products (Ghosal, Indira & Bhattacharya, 2010) EC composition influenced wettability of jussara pulp microparticles, which was inversely associated with hygroscopicity Jussara pulp microparticles produced solely with OSA starch (run #1) showed the slowest wettability (267 s), approximately 4-fold slower that the average wettability of the remainder microparticles (ranging from 41 s to 91 s) (Table 3) Thus, we considered that these microparticles produced with OSA starch as the only coating material were unsuitable for producing food colorants This behavior may be justified by the lipophilic character of OSA starch, as previously mentioned for hygroscopicity Fernandes, Borges and Brotel (2014) also observed higher wettability values for rosemary oil microencapsulated with OSA starch (254 s) when compared to a mixture of inulin and gum arabic (93 s) The fastest wettability of runs #28 (43 s) and #31 (41 s) suggest that using high core to EC ratios improve the instantaneous characteristics of microparticles, which may be explained by their agglomeration (Fig 3), as well as larger mean particle size (Table 2), leading to more spaces between particles and thus easing the penetration of water into the pores (Ghosal et al., 2010) The inverse association between wettability and mean particle size was also observed for blackberry microparticles produced with maltodextrin (82 s/49 m), gum arabic (134 s/11 m) and their mixture (116 s/28 m) (Ferrari et al., 2012) Morphological characterization of selected jussara pulp microparticles (Fig 3) showed continuous surfaces with no fis- 507 Table Water solubility, hygroscopicity and wettability of jussara pulp microparticles selected from the experimental design of Table 1.a Run Solubilityb (%) Hygroscopicityc (%) Wettabilityd (s) 18 19 24 28 31 83.1 ± 4.2 76.8 ± 1.21 77.6 ± 2.91 82.0 ± 0.71 80.8 ± 0.11 85.0 ± 1.91 80.7 ± 0.11 10.7 ± 0.4 12.0 ± 0.32,3 11.5 ± 0.01,2 14.1 ± 0.14 13.7 ± 0.03,4 12.6 ± 0.63 14.3 ± 0.04 267 ± 154 79 ± 13 91 ± 113 77 ± 12,3 86 ± 73 43 ± 11,2 41 ± 41 1 a Results expressed as mean ± SD for triplicates Different superscript Arabic numerals in each column indicate significant difference between jussara pulp microparticles (Oneway ANOVA followed by Tukey test, p < 0.05) b Percent mass of microparticles solubilized in water c Mass gain of microparticles stored for days at 75% relative humidity and room temperature d Time required for microparticles to be visually completely wet sures, cracks or interruptions, which is essential to ensure low gas permeability and good protection of anthocyanins In general, microparticles’ showed spherical shape and various sizes independently of EC composition and core to EC ratio The surface of microparticles produced with 0.5 and 1.0 core to EC ratios presented roughness, showing dents and slight invaginations (Fig 3) The observed surface roughness is characteristic of microparticles produced by spray drying at low drying temperatures In this case, the microparticles surface remains moist and supple during the drying process, so the particle deflates and shrivels as it cools down (Tonon et al., 2008) Microparticles produced with 2.0 core to EC ratio and high proportions of inulin showed a smoother surface (Fig 3), possibly due to the higher molecular flexibility of this carbohydrate, which shows a multitude of possible conformations (Mensink, Frijlink, Maarschalk & Hinrichs, 2015) Agglomerated jussara pulp microparticles were visible in all selected conditions, being more loosely bound for those produced with 0.5 and 1.0 core to EC ratios Formation of link bridges between microparticles and more intense agglomeration was observed in microparticles produced with 2.0 core to EC ratio, especially when inulin was used (runs #24 and #28) (Fig 3) From each core to EC ratio, we selected a single condition run for the stability test at 50 ◦ C Run #9 was selected over run #1 based on their wettability values The choice between runs #18 and #19 was based on their hygroscopicity, anthocyanins content and L* value, all of them favoring the former Among microparticles produced at 2.0 core to EC ratio, run #28 was excluded due to their intense agglomeration Finally, the choice between runs #24 and #31 was based on their EC composition, favoring the former due to its higher proportion of inulin, which shows dietary fiber actions, among other health related benefits (Ranawana, 2008) 3.3 Stability of jussara pulp microparticles during storage at 50 ◦ C Stability of color, anthocyanins and antioxidant activity (FRAP and TEAC) of the jussara pulp microparticles stored at 50 ◦ C either unprotected and protected from light for 38 days are shown in Fig As a general trend, jussara microparticles color became slightly lighter and less red during storage, although their anthocyanin contents and FRAP values did not change AA measured by TEAC assay, on the other hand, decreased during storage, especially in early stages Moreover, light protected and unprotected jussara microparticles showed the same behavior for all the variables investigated These results suggest that, in general, EC were effective in protecting jussara pulp from degrading conditions The color of all three microparticles changed after 38 days of storage at 50 ◦ C, independently of light exposure (Table 4) Based on 508 E.C.Q Lacerda et al / Carbohydrate Polymers 151 (2016) 500–510 Fig Scanning electron micrographs (2000 × magnification) of jussara pulp microparticles produced with mixtures of OSA starch, inulin and maltodextrin at different core to encapsulating carbohydrates (EC) ratios, according to the experimental design shown in Table Table Instrumental color (L*a*b*), total color differences ( E*), anthocyanins contents and antioxidant activity (FRAP and TEAC assays) at t = and t = 38 days in jussara pulp microparticles produced with OSA starch (OSA), inulin (I) and maltodextrin (M) at different core to encapsulating carbohydrates (EC) ratios, stored at 50 ◦ C unprotected and protected from light.a Component 0.5 core to EC ratio1/6 OSA, 1/6 I and 2/3 M 1.0 core to EC ratio 2/3 OSA, 1/6 I and 1/6 M 2.0 core to EC ratio t=0 t=0 t=0 t = 38 days Unprotected L* a* b* E* Anthocyanins (mg/g) FRAP (mol Fe+2 /g) TEAC(mol Trolox/g) 35.8 22.9 0.7 11.6 696.9 382.8 * 36.0 (1%) 21.9* (−5%) 0.5* (−31%) 1.1 12.8 658.4 263.9* (−31%) Protected *# 34.7 (−3%) 21.7*# (−5%) 0.08*# (−89%) 1.8# 12.3 676.7 259.0* (−32%) t = 38 days Unprotected 28.9 20.6 1.6 10.8 1290.1 586.9 * 30.9 (7%) 20.9* (1%) 1.0* (−37%) 2.1 11.4 1257.8 384.1* (−35%) Protected *# 29.9 (3%) 20.7*# (1%) 1.0*# (−34%) 1.1# 11.9* (10%) 1388.0 406.9* (−31%) t = 38 days Unprotected 23.6 16.1 1.6 13.8 1568.1 882.9 1/1 I * 26.1 (11%) 15.4* (−4%) 0.8* (−47%) 2.7 13.7 1233.0* (−21%) 340.1* (−61%) Protected 24.7*# (5%) 14.2*# (−12%) 0.8* (−47%) 2.3# 14.1 1381.1* (−11%) 357.9* (−69%) a Means at t = 38 days with a superscript asterisk are significantly different from their corresponding sample at t = Relative change in relation to t = is shown in parenthesis Means at t = 38 days protected from light with a superscript hashtag are significantly different from the corresponding unprotected sample (ANOVA, Tukey’s post-test; p < 0.05) Differences for E* between unprotected and protected sample were evaluated by unpaired t test (p < 0.05) All results presented a coefficient of variation lower than 8.0% E.C.Q Lacerda et al / Carbohydrate Polymers 151 (2016) 500–510 509 Fig Stability of color (L*, a* and E*), anthocyanins contents and antioxidant activity (FRAP and TEAC assays) of jussara pulp microparticles produced with mixtures of OSA starch, inulin and maltodextrin at core to encapsulating carbohydrates ratios of 0.5 ( ), 1.0 ( ) and 2.0 ( ) stored at 50 ◦ C unprotected (dashed line and open symbols) and protected (full line and closed symbols) from light for 38 days E* values, which correspond to the total color difference between initial and final time points, microparticles showed slight color differences during storage at 50 ◦ C, as well as due to light exposure (Table 4) Microparticles produced with 1.0 and 2.0 core to EC ratios which were protected from light showed lower E* than those exposed to light On the other hand, microparticles produced with 0.5 core showed the opposite behavior, that is higher E* values when protected from light The highest losses in original total color ( E* = 2.5, on average) were observed for microparticles produced at 2.0 core to EC ratio According to Óbon, Castellar, Alacid and López (2009), 0.0< E*